Four kinds of knowing, and the mantra named at last

Vidyā is from √vid, to know, the root behind veda and, at a great distance, behind English wit and Latin video. In ordinary Sanskrit it means knowledge, a science, a branch of learning. In tantric usage it acquires a narrower technical sense: a vidyā is a mantra of a goddess, as against a mantra proper, which belongs to a god. Four of the names in this verse are therefore doing double duty, and the reader who hears only the general sense of learning will miss half the verse.

Ātmavidyā is the older term. The Chāndogya Upaniṣad stages the classic scene: Nārada comes to Sanatkumāra having mastered the Vedas, grammar, mathematics, astronomy, logic, the science of ghosts and the science of snakes, and confesses that he knows formulas but does not know the self, and that he has heard the knower of the self crosses over sorrow. The Muṇḍaka draws the same line more sharply, dividing all knowledge into a lower kind, which includes the Vedas themselves, and a higher kind by which the imperishable is reached. To open the verse with ātmavidyā is to place this tradition on the far side of that division before saying anything about mantras at all.

Mahāvidyā can be read simply as the great knowledge, and most reciters take it that way. It is also, in the wider Śākta literature, the technical name of a group: the ten Mahāvidyās, a set of goddesses beginning with Kālī and Tārā and including Tripurasundarī, who is Lalitā herself, along with figures as strange as the headless Chinnamastā and the widowed Dhūmāvatī. The lists vary between sources and the group is attested later than some of its members, so it would be wrong to read the name as a straightforward reference to a fixed canon. What can be said is that the word was available in both senses when this hymn was composed, and that both are current among those who recite it.

Śrīvidyā is the name of the school. Śrī covers radiance, prosperity, well-being and the goddess Lakṣmī, and the tradition is named for it because the formula is held to confer all of these along with liberation, which is the claim distinguishing this system from the sixty-four partial tantras. The vidyā in question is the pañcadaśākṣarī, the fifteen-syllabled mantra, and the following names describe it.

Kāmasevitā places the god of desire in the posture of a servant. Two strands of story converge here. Kāma is the archer with the flower bow whom Śiva burnt to ash with the third eye for disturbing his meditation, and in the Lalitā narrative it is from that ash, worked into a figure and given life, that the demon Bhaṇḍāsura arises — so the goddess who eventually destroys Bhaṇḍa is dealing with the consequences of Kāma's incineration. More directly, Kāma under the name Manmatha is one of the twelve traditional propounders of this vidyā, and the lineage descending from him, whose formula begins with the syllable ka, is the kādi school, the dominant one. Kāmasevitā is thus not only a picture but a statement about transmission: the tradition holds that desire itself received this teaching and passed it on.

The fifteen-syllabled mantra is divided into three segments called kūṭas, peaks, of five, six and four syllables. They are named vāgbhava, the origin of speech; kāmarāja, the king of desire; and śakti, power. Each ends with the syllable hrīṃ, which the texts call the hṛllekhā. Trikūṭā, she of three peaks, is the goddess described as this structure, and the three segments are correlated in the commentarial literature with almost every triad the tradition uses — the three powers of will, knowledge and action; waking, dream and deep sleep; the three cities that give Tripurā her name.

Śrīṣoḍaśākṣarīvidyā names the further form in which a sixteenth syllable is added to the fifteen. The additional syllable is treated as reserved for initiation; it is most often reported as śrīṃ, but the sources are not uniform and the tradition has never printed it as freely as it prints the rest. The hymn's own procedure is instructive: it names the mantra without giving it. Kāmakoṭikā closes the verse on a word with two senses. Koṭi means the tip or extreme point of anything, so that the name says she is where desire terminates; koṭi also means ten million, which yields the alternative reading, she of countless Kāmas, and the module records both. The word is also the name of the Kāmakoṭi seat at Kāñcī, a town whose presiding goddess Kāmākṣī has long been identified with Lalitā, and reciters in the south hear the place-name in the name whether or not the composer intended it.

Four names in this verse end in the same three syllables, and that is the first practical problem. Vidyā contains the conjunct dyā, a voiced dental d followed immediately by the palatal glide y: the tongue starts at the teeth and moves back, and there must be no vowel between the two. Say it four times in the shape the verse gives it — ātmavidyā, mahāvidyā, śrīvidyā, then the long one — and let the increasing length of the qualifier keep them apart in your memory, since the endings will not.

Śrī is worth a minute on its own, because it is the most frequently mispronounced syllable in the tradition. It is palatal ś, then r, then long ī, all in one syllable. It is not shree with an English r, and it is not the retroflex ṣ. Contrast it directly with the next name, where ṣoḍaśa opens with the retroflex ṣ, the tongue curled back, and then returns to palatal ś two syllables later. Śrīṣoḍaśākṣarīvidyā puts all of that in a row and then adds kṣ: ś-r-ī, ṣ-o, ḍ-a, ś-ā, k-ṣ-a, r-ī, v-i, d-y-ā. Sanskrit has three sibilants and this single name uses two of them four times, alternating. Nothing else in the hymn drills the distinction so efficiently.

Trikūṭā and kāmakoṭikā both turn on the retroflex ṭ, made with the tongue tip curled up against the roof of the mouth, quite unlike the dental t of ātma. The ū of trikūṭā and the o of koṭikā are both long. Kāmakoṭikā has four syllables beginning with k and only the vowels to distinguish them, so keep ā, o, i and ā clearly separated or the word turns to porridge.

Metrically the verse is regular, eight syllables to the quarter, but the distribution of names across it is uneven: four names fill the first line and three the second, with the eight-syllable śrīṣoḍaśākṣarīvidyā taking a whole quarter by itself. The pace necessarily slows in the second line.

In the archanā all seven take the same ending, since every one is an ā-stem: ātmavidyāyai, mahāvidyāyai, śrīvidyāyai, kāmasevitāyai, śrīṣoḍaśākṣarīvidyāyai, trikūṭāyai, kāmakoṭikāyai. After the mixed declensions of the preceding verses this uniformity is a relief, and it makes the verse a good one to use when teaching the dative to someone learning the offering.

What the verse asks is a change in what the reciter takes himself to be doing. Up to this point the hymn has been describing a deity. Here it describes the practice, and it names it as a form of her — so that an initiate reciting śrīvidyā is naming both the goddess and the thing he was given at initiation, and the two names are the same name. For someone reciting without initiation, which is the ordinary case and a perfectly proper one, the verse is still doing something: it is stating plainly that ātmavidyā and śrīvidyā are not two subjects, and that a mantra of sixteen syllables and the knowledge of the self are being claimed as one object under two descriptions. That claim is either the centre of the tradition or its most extravagant assertion, and this is the verse where it is made without hedging.